Alex Gabel

Physics. Film. AI. et al.

[Alex Gabel @gabel_alex]

Next time, curvature! Notes will be available online soon📚

An Introduction to Differential Manifolds (Barden & Thomas) and Introduction to Smooth Manifolds (Lee) were used in preparation 📖

@deepthinkhour events are supported by @Ellis_Amsterdam and hosted at @UvA_Amsterdam

[Alex Gabel @gabel_alex]

Last week, we wrapped up the first part (Foundations) of the "Differential Geometry for Deep Learning" series! 📐

Thanks to all the attendees and of course @davidmknigge, @samuelepapa_, and @RValperga for making these lectures possible!

[David M. Knigge @davidmknigge]

✨Good news! The lovely @gabel_alex is back to give the second lecture of his tutorial series on "Differential Geometry for Deep Learning" at @deepthinkhour.

In this follow-up Alex will treat a bunch of new topics, as well as showcasing a number of practical applications 👇

[ELLIS Amsterdam @Ellis_Amsterdam]

💥 We are excited that @davidwromero (@nvidia) will talk about 'Beyond Transformers: Exploring Subquadratic Long-Context Architectures' at the upcoming Deep Thinking Hour Tutorial!

📅 Thu, April 11th

⏰️09:00 - 11:00

📍L1.01 of @Lab42UvA

Come and deep think with us! 🏍

[David M. Knigge @davidmknigge]

🗓️ Reminder! @phillip_lippe's tutorial on training massive deep learning models is happening this afternoon at 17:00 CET in C0.110 SP904.

Also live-streamed over zoom. For info and links:
https://t.co/exhuB0hdK0 ✨

[Alex Gabel @gabel_alex]

We also propose using a suite of 25 PDEs labeled by their non-canonical symmetries for experimenting with neural symmetry detection beyond rotation, scaling and translation! (n/n)

And, of course, big thanks to my supervisors Stratis Gavves and Rick Quax here at @UvA_Amsterdam!

[Alex Gabel @gabel_alex]

Our work got published in Machine Learning: Sci & Tech! 🎉 @egavves @MLSTjournal

We propose a method for encoding complicated symmetries pervasive in scientific data, apply it to neural symmetry detection AND argue for more diverse PDE datasets! (1/n)

https://t.co/lUuUn1Z6al

[Alex Gabel @gabel_alex]

In a straightforward experiment, we show how symmetry detection can be done for PDE data using this encoding. In some cases, parts of correct symmetry transformations were identified when applied to unseen dynamics 👉 Exciting possibility for further work! (3/n)

[Phillip Lippe @phillip_lippe]

Do you want to train massive deep learning models with ease? Our 10 new tutorial notebooks of our popular UvA DL course show you how, implementing data, pipeline and tensor parallelism (and more) from scratch in JAX+Flax! 🚀🚀
Check them out here: https://t.co/7AtaB7j8l9
🧵 1/11

[Machine Learning: Science and Technology @MLSTjournal]

Great new work by @gabel_alex Rick Quax and @egavves @UvA_CSL @UvA_IAS @UvA_Amsterdam - 'Data-driven Lie point #symmetry detection for continuous #dynamicalsystems' -https://t.co/HOvT6cjM5C #machinelearning #complexity #datascience #compphys #AI #neuralnetworks #PDEs #informatics

[David M. Knigge @davidmknigge]

The @UvA_Amsterdam and @ELLISforEurope are starting a masterclass series where experts explain all about their favourite topics in and around Machine Learning.

Join us for the first on March 6th; "Differential Geometry for Deep Learning" masterclass by @gabel_alex. ✨

👇

[Alex Gabel @gabel_alex]

@MiltosKofinas Congratulations! 🥳

[Hazel Doughty @doughty_hazel]

I have two #PhD openings in #ComputerVision on 'Detailed Video Understanding' at @LIACS
@UniLeiden.

Apply before 10th November. More info 👇
https://t.co/zMN6yEHywQ

[Alex Gabel @gabel_alex]

Check out our poster on Neural Symmetry Detection this morning at TAG-ML! (Room 317B at 11.10am) @TAGinDS #ICML2023

Work with: @klein_vick @RValperga @kn_webster Jeroen Lamb @RickQuax & @egavves

https://t.co/FaeQB27yJQ

[David W. Romero @davidwromero]

Tired of tuning neural architectures by hand every time you have a new dataset, modality?

Meet 𝐃𝐍𝐀𝐫𝐜𝐡, a method that jointly learns:
• The size of all conv kernels
• The width of all layers
• Downsampling layers &
• The depth of the nets
all by backpropagation! 🔥

[Alex Gabel @gabel_alex]

I’m at ICML in Hawaii this week! 🏄‍♂️ 🌸 ☀️

Let me know if you’re around and want to meet up ☕️

We’re presenting our work on Neural Symmetry Detection on Friday at the TAG-ML poster session 🤖 📐

#ICML2023

Link: https://t.co/FaeQB27yJQ

[Max van Spengler @MvanSpengler]

Excited to introduce HypLL, our brand-new hyperbolic learning library developed together with @phippli and @PascalMettes!

1/4

[Fabian Dablander (@fdabl.bsky.social) @fdabl]

In a new blog post, I summarize and reflect on some of the themes of our Simulation-based Science colloquium, which we recently concluded @UvA_IAS.

I learned a lot from our excellent speakers, and I am excited to see what future iterations will hold!

https://t.co/QKcB6AmeoL

[Alex Gabel @gabel_alex]

@egavves A potential research sibling!